2 Location and Geological setting
5.1 Chemical effects and styles of alteration
(Percent) (Percent) (Pixels) (Pixels)
Zeolithe 76.92 100 10/13 10/10
ORCO alteration 60.24 83.33 50/83 50/60
CHAV scree higher‐T 59.21 84.91 45/76 45/53
CHAV high‐T silicification 87.93 100 51/58 51/51
CHAV scree 83.33 25.64 10/12 10/39
Iron oxides 86.96 100 20/23 20/20
Weathered Ignimbrite 90.63 65.91 29/32 29/44
CCAR scree 84 89.36 42/50 42/47
CCAR scree with clay minerals 100 37.04 10/10 10/27
Soil/Regolith (unclassified) 100 79.26 172/172 172/217
Lava 77.44 100 127/164 127/127
Ignimbrite 61.29 86.36 19/31 19/22
Overall Accuracy (585/724) 80.80%
Kappa Coefficient 0.7755
In summary, the classification algorithm distinguishes between highly silicified regions in the volcanic core (“CHAV high-T silicification”), areas with argillic alteration, silicification and iron oxides to variable degrees (“ORCO alteration”, “CCAR scree”, “CCAR scree with clay minerals”, “CHAV scree”), areas with dominating iron oxides (“iron oxides”), ignimbrite outcrops, lava and “zeolite alteration”. The differences between the “scree” classes lies in the amount of iron oxides and clay minerals and the amount of silicification. This can be seen in different absorption depth for different samples, in the geochemical data and in Fig. 4.
Therefore, on each epithermally altered volcano, we find all scree classes represented. This shows the similarity style of of alteration on all volcanic centers and the gradual change from silicification to areas with increasing amount of clay minerals and iron oxides.
5 Discussion
5.1 Chemical effects and styles of alteration
Our analysis shows that the results of the spectral analysis of our ground-truth areas in the field, laboratory spectral measurements on sampled material, as well as geochemical analysis results are mutually consistent (Table 4). Alteration styles were identified and
Manuscript II ‐ Mapping patterns of mineral alteration in volcanic terrains using ASTER data and field spectrometry in Southern Peru
Chapter 3 63
mapped successfully by combining these approaches in interpreting the ASTER spectral data. We will now link these different styles of alteration to bulk geochemical changes in major and trace element composition of the altered volcanic rocks.
Hydrothermal alteration results in changes of major element contents that becomes apparent in a plot of K2O and the chemical index of alteration (CIA) versus SiO2 (Fig. 15, 16). While magmatic differentiation should lead to an increase in K2O with an increasing SiO2 content, hydrothermal fluids leach potassium preferentially to other oxides (Al2O3) that remain or can even be actively enriched (SiO2) in the process. At the same time the CIA increases with mobilisation and decreasing contents of Na, Ca and K. Only under extremely acid or alkaline conditions Al becomes also mobile (Hay and Sheppard, 2001) and under these conditions the CIA loses its measure for alteration. Andesite from Cerro Orconccocha and Cerro Palla Palla show high degrees of leaching and silicification, where we observe silicification up to almost 100% SiO2 ( Fig. 15).
Fig. 15: Element variation plot for SiO2 (wt-%) versus K2O (wt-%). Arrows show trends defined by differentiation, iron enrichment, silicification and hydrothermal leaching. Fresh Ignimbrite and lava samples collected in the same area are plotted together with alteration samples. Samples from Cerro Palla Palla and from Cerro Orconccocha show strong leaching and silicification while samples from Lomada Atansa and Cerro Carhuarazo are closer to the differentiation trend or show iron enrichment.
Manuscript II ‐ Mapping patterns of mineral alteration in volcanic terrains using ASTER data and field spectrometry in Southern Peru
Chapter 3 64
Fig. 16: This diagram shows the Chemical index of alteration Nesbitt and Young, 1982) plotted against SiO2. Fresh samples are shown in yellow for comparison. Different processes are shown with arrows. Hydrothermal alteration first leaches the rock of alkaline elements and active SiO2 enrichment occurs. Under extremely acid or alkaline conditions aluminium becomes mobile and the CIA decreases.
According to Corbett and Leach (1997), the formation of alteration minerals is determined by a number of variables that can be grouped into seven main categories (Browne, 1978):
temperature, fluid chemistry, concentrations, host rock composition, kinetics of reactions, duration of activity or degree of equilibrium and permeability. SiO2 contents of unaltered volcanic rocks in our study area range from about 53 weight-% to 73 weight-% (29 lava and ignimbrite pumice samples). Temperature and fluid chemistry have the strongest influence on the style of hydrothermal alteration, although all of the stated factors are somewhat interdependent. In a diagram that was derived from a data compilation by Corbett and Leach (1997), fluid pH and temperature were used to show stability ranges for common hydrothermal alteration minerals. Although the stability ranges can change due to other factors, the diagram in his study serves as a good approximation for stability ranges here.
Characterizing our own results accordingly, we observe mostly silica group minerals grading to kaolinite/illite group minerals. Highest temperatures and lowest pH occur on top of Cerro Palla Palla (spectral class “CHAV-high-T silicification”) and on Cerro Orconcchocha (spectral class “ORCO alteration”). Silica group minerals indicate a fluid pH lower than 2 (Stoffregen, 1987). We attribute the very high content of amorphous silica to the fact that hydrothermal fluids must have been highly acidic and oversaturated with respect to SiO2. Local hydrothermal alteration recognized in the Incapacha project, a detailed study of Cerro Incapacha and surroundings in the SW of our area (Guevara, 2001) was described as silicic
Manuscript II ‐ Mapping patterns of mineral alteration in volcanic terrains using ASTER data and field spectrometry in Southern Peru
Chapter 3 65
to advanced argillic and confirms our findings. With increasing pH and decreasing temperature, alunite, and clay minerals: first kaolinite at pH 3-4, T < 200-150°C and then illite/smectites (pH 4-5, T < 100-200°C) become stable mineral phases (Corbett and Leach, 1997). Using spectral classification on ASTER data, we can visualize these changing conditions and stability ranges.
SVM classification identifies zones of silicification that are critical for exploration (Hedenquist et al. 2000) and that are not easily captured by simple and often-used ratio-images. The iron absorption feature normally cannot be seen clearly in ASTER data but the deep trough of the feature still affects ASTER band 3 and band 1 is pulled down by the iron absorption as well.
Therefore, a high content of iron can be distinguished because the ASTER spectrum has high ratios for band 2/ band 1 and band 3/ band 2 (“iron oxides” in Fig. 4) and is captured by the classification algorithm. The five bands in the SWIR centered around 2 µm to 2.5 µm catch important absorption features of clay minerals. Zeolite absorption features are centered at 1.4 µm and 1.9 µm and cannot be detected by ASTER due to atmospheric absorption.
Nevertheless, the form of the spectrum from bands 3 to 9 is very pronounced with low reflection of band 3 and 9 and high reflection values for bands 4, 6 and 7 (Fig. 4). This allows distinction of this alteration class in our scenes and is in accordance with pure zeolite spectra from USGS spectral library resampled to ASTER spectral resolution.
Results from geochemical, spectral analysis and classification allow the following observations: 1) We identify areas (stratovolcanos) that show hydrothermal alteration typical for high temperatures and low pH values (< 2) and areas with higher pH values (4-5) and argillic alteration. The zonation around a silicic high-T alteration core is striking. 2) There are large areas covered with ignimbrites that are surrounded by the altered volcanic structures.
3) The Andamarca formation is cut by minor faults with zeolitic alteration (pH neutral to alkaline for zeolite stability).
Table 4: Comparison of geochemical and spectral results
Results from ASD spectral analysis, geochemical and mineral sensitive methods (GMS) for the different alteration classes*
Alteration mineral group** Zeolithe alteration weathered ignimbrite CCAR scree CCAR scree with clay minerals CHAV scree
Silica group minerals ASD x x x x
Alunite group minerals ASD
Kaolin group minerals ASD x
Illite group minerals ASD x x x x
Calc‐silicate group ASD x
Silica group minerals GMS x x x
Alunite group minerals GMS x
Kaolin group minerals GMS x
Illite group minerals GMS x x
Calc‐silicate group GMS x
* Samples used to define the class are given in tabn.a.: samples not analyzed
Alteration mineral group** CHAV scree highter‐CHAV high‐T silicificationORCO alteratIron oxides
Silica group minerals ASD x x x x
Alunite group minerals ASD
Kaolin group minerals ASD x x
Illite group minerals ASD x x
Calc‐silicate group ASD
Silica group minerals GMS x x x x
Alunite group minerals GMS
Kaolin group minerals GMS x
Illite group minerals GMS x
Calc‐silicate group GMS
** mineral groups after Corbett&Leach, 1997
Manuscript II ‐ Mapping patterns of mineral alteration in volcanic terrains using ASTER data and field spectrometry in Southern Peru
Chapter 3 67